BMC BioMed Central

Methodology article Open Access Displacement affinity chromatography of protein phosphatase one (PP1) complexes Greg BG Moorhead*1, Laura Trinkle-Mulcahy2, Mhairi Nimick1, Veerle De Wever1, David G Campbell3, Robert Gourlay3, Yun Wah Lam2 and Angus I Lamond2

Address: 1Department of Biological Sciences, University of Calgary, 2500 University Dr. N.W. Calgary, AB T2N 1N4, Canada, 2Wellcome Trust Biocentre, MSI/WTB Complex, University of Dundee, Dundee, DD1 5EH, UK and 3MRC Protein Phosphorylation Unit, School of Life Sciences, University of Dundee, Dundee, Scotland DD1 5EH, UK Email: Greg BG Moorhead* - [email protected]; Laura Trinkle-Mulcahy - [email protected]; Mhairi Nimick - [email protected]; Veerle De Wever - [email protected]; David G Campbell - [email protected]; Robert Gourlay - [email protected]; Yun Wah Lam - [email protected]; Angus I Lamond - [email protected] * Corresponding author

Published: 10 November 2008 Received: 8 May 2008 Accepted: 10 November 2008 BMC Biochemistry 2008, 9:28 doi:10.1186/1471-2091-9-28 This article is available from: http://www.biomedcentral.com/1471-2091/9/28 © 2008 Moorhead et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract Background: Protein phosphatase one (PP1) is a ubiquitously expressed, highly conserved protein phosphatase that dephosphorylates target protein serine and threonine residues. PP1 is localized to its site of action by interacting with targeting or regulatory proteins, a majority of which contains a primary docking site referred to as the RVXF/W motif. Results: We demonstrate that a peptide based on the RVXF/W motif can effectively displace PP1 bound proteins from PP1 retained on the phosphatase affinity matrix microcystin-Sepharose. Subsequent co-immunoprecipitation experiments confirmed that each identified binding protein was either a direct PP1 interactor or was in a complex that contains PP1. Our results have linked PP1 to numerous new nuclear functions and proteins, including Ki-67, Rif-1, topoisomerase IIα, several nuclear , NUP153 and the TRRAP complex. Conclusion: This modification of the microcystin-Sepharose technique offers an effective means of purifying novel PP1 regulatory subunits and associated proteins and provides a simple method to uncover a link between PP1 and additional cellular processes.

Background groups based on protein sequence, catalytic signature and The phosphorylation of proteins is one of the most prev- substrate preference [3-5]. The action of protein phos- alent covalent modifications known, affecting essentially phatases is tightly controlled with cellular targeting being every aspect of cellular function [1,2]. The protein kinases an important means of regulation. Most phospho-serine and phosphatases responsible are highly conserved across and threonine dephosphorylation can be attributed to the species and, with few exceptions, the kinases belong to PPM family and the more diverse PPP family, which one large gene family while the phosphatase complement includes PP1, PP2A, PP2B, and PP4 through to PP7. PP1 is more complex and can be divided into three broad is thought to not exist as a free catalytic subunit in the cell,

Page 1 of 10 (page number not for citation purposes) BMC Biochemistry 2008, 9:28 http://www.biomedcentral.com/1471-2091/9/28

but to reside in complexes with a large array of targeting addition to in vivo targeting, disrupting PP1-regulatory or regulatory subunits that define its function. Numerous subunit interactions with a peptide would be an effective PP1 docking proteins have been identified, but they most means to aid in identifying proteins in PP1 complexes and likely represent only a small fraction of the total number thus uncover new cellular processes regulated by this pro- in the cell. tein phosphatase.

The microcystins are a group of cyclic peptides that bind Results and Discussion with remarkable specificity and affinity to the type one, 2A We initiated our PP1 peptide displacement study selecting and several recently identified protein phosphatases of the RVXF/W containing peptides from the PP1 targeting sub- PPP family (e.g. PP4, PP6). Microcystin covalently cou- units NIPP1 [25,30] and ZAP (ZAP3) [31,32]. They were ples to a conserved cysteine residue of PPP family mem- synthesized and tested for their ability to displace PP1 bers through its methyl-dehydroalanine residue [6,7]. binding proteins from complexes retained on the micro- Nishiwaki et al [8] first used Microcystin-Sepharose to cystin matrix. In brief, we isolated rat liver nuclei, purify PP2A. We exploited a different synthetic approach extracted proteins and incubated extracts with Microcys- whereby the carbon-carbon double bond of methyl-dehy- tin-Sepharose to bind the microcystin-sensitive protein droalanine in microcystin couples the latter to ami- phosphatases [31]. After extensive column washing, we noethanethiol, which is then linked to a Sepharose bead. incubated the matrix with the NIPP1 or ZAP peptides This generates a high affinity binding matrix for the (RPKRKRKNSRVTFSEDDEII and GKKRVRWADLE, microcystin-sensitive protein phosphatases that does not respectively) to selectively displace proteins retained on covalently couple the phosphatase [6]. Microcystin- the matrix through PP1. This implies that PP1 itself and Sepharose has proved to be a powerful tool to purify these other microcystin-sensitive phosphatase complexes protein phosphatases and their associated regulatory sub- should be retained on the matrix and can subsequently be units from a variety of tissues and cell types [9-12]. eluted with the chaotrophic agent sodium thiocyanate (NaSCN), which will also displace other bound proteins With only a few characterized exceptions, PP1 interacting [9] including some that may be retained by binding non- proteins bind PP1 through their primary docking specifically to the Sepharose bead. Blotting for PP1 and sequence called the RVXF/W motif [13]. Their molecular the PP2A regulatory subunit PR65 show this to be true interaction with PP1 has been visualized via PP1-peptide (Figure 1a and 1b) with no retention of these phosphatase and PP1 regulatory subunit structures [14,15]. It has also subunits on a control matrix coupled with Tris alone. The emerged that additional or secondary interaction sites NIPP1 peptide (RPKRKRKNSRVTFSEDDEII) readily dis- often play a role in binding PP1 and likely contribute to placed PP1 regulatory subunits (data not shown), yet we PP1 isoform specificity recognition, substrate docking and chose to continue working with the ZAP peptide modulation of PP1 activity [13,15-20]. Based on a compi- (GKKRVRWADLE, later also referred to as RVRW peptide) lation of demonstrated RVXF/W interaction motifs [21- because it is small, readily soluble and an excellent match 23], the panning of a random peptide library [24], and to the optimal PP1 binding peptide discovered through mutagenesis and modeling studies [23] we noted prefer- panning a random peptide library [24]. Indeed, the ZAP ences for particular amino acids within and adjacent to peptide RVRW motif and the additional C-terminal the RVXF/W motif. This led us to speculate that a PP1 amino acids ADL and N-terminal basic amino acids (KK) interaction motif peptide, based on this comparison, were most frequently obtained in the random peptide could be a unique means to specifically disrupt PP1-tar- library screen (Figure 2 in [24]). We thus predict that the getting or regulatory subunit interactions [14,25,26]. ZAP sequence GKKRVRWADLE is the most suitable, high Slight variation in the RVXF/W-motif combined with the affinity peptide that will compete and displace most other now recognized additional, secondary PP1 interaction proteins docked to PP1 through variations of the RVXF/ sites provide sufficient interaction specificity which may W-motif. We acknowledge that even with an 'optimal allow the development of drugs or peptide mimetics to binding peptide' there are certainly other PP1 interacting abolish specific PP1 binding protein interactions in vivo. proteins on the matrix that may not be displaced because The idea of targeting protein-protein interaction domains of completely novel PP1 interaction sites and/or addi- with drugs has historically not been favored by the phar- tional interaction sites outside of the RVXF/W-motif [20] maceutical industry yet, due to improved understanding that could maintain the interaction with PP1 even if the of the underlying molecular mechanisms, it is now a con- RVXF/W site is displaced. Thus, the PP1 binding partners cept that is growing in popularity [27,28]. This idea has uncovered here still most likely only represent a sub-pop- been explored with PKA anchoring proteins (AKAPs) ulation of the total nuclear PP1 binding partners. where optimal RI and RII subunit binding peptides were derived from parent peptides and used to target PKA in RVRW peptide eluted samples were blotted for the known vivo and displace it from its normal anchoring site [29]. In nuclear PP1 binding proteins ZAP, p99 (PNUTS) and

Page 2 of 10 (page number not for citation purposes) BMC Biochemistry 2008, 9:28 http://www.biomedcentral.com/1471-2091/9/28

Figure 1 A peptide based on the PP1 interaction motif RVXF/W displaces PP1 regulatory subunits from the affinity matrix microcystin-Sepharose. Proteins extracted from isolated rat liver nuclei were incubated with the affinity (MC) or control (Con) matrix (Tris coupled). Proteins were displaced with the RPKRKRKNSRVTFSEDDEII peptide (a), while in (b), elution was performed with either RPKRKRKNSRVTFSEDDEII or GKKRVRWADLE prior to the 3 M NaSCN elution. The control matrix is only used in the panel (a) experiment while microcystin-Sepharose is used in all others. After concentration, samples were run on 10% SDS-PAGE, blotted to nitrocellulose and membranes were probed with anti-PR65 and PP1 antibod- ies. In panel (c) the membrane was probed with anti-ZAP, p99 or NIPP1 antibodies [31]. To test for the salt dependence of peptide displacement, columns were eluted with peptide plus or minus NaCl as indicated and samples blotted for NIPP1 (d). To determine optimal peptide concentration for displacement from the column, the protein loaded beads were divided into 3 equal parts and eluted with 0.1, 0.5 and 2 mM peptide (e). To test for the specificity of the peptide displacement, the beads were divided into 3 equal parts and eluted with either GKKRVRWADLE, the GKKRVRWADLE peptide with the key inter- acting residues changed to A (GKKRARAADLE), or a scrambled version of the GKKRVRWADLE peptide (KLRGEVAK- DWR) and blotted for ZAP, p99 and NIPP1. Glycogen particles were isolated from rabbit skeletal muscle and PP1GM bound to the microcystin matrix followed by peptide elution first with the GKKRARAADLE peptide, then GKKRVRWADLE, followed by 3 M NaSCN.

Page 3 of 10 (page number not for citation purposes) BMC Biochemistry 2008, 9:28 http://www.biomedcentral.com/1471-2091/9/28

NIPP1 (Figure 1c). This revealed that these proteins were Furthermore, peptides may not be detected during mass readily displaced from the matrix by the peptide with spectrometry in the presence of more abundant peptides. essentially no further regulatory subunit being eluted with Using an extract prepared from isolated rat liver nuclei we NaSCN. A direct comparison of elution conditions with could show that GKKRVRWADLE peptide displacement the GKKRVRWADLE peptide showed that displacement gave a banding pattern on SDS-PAGE that was clearly dis- was much more effective in a higher ionic strength buffer tinct from the subsequent NaSCN elution [Additional file (Figure 1d). Having established that the GKKRVRWADLE 1]. Individual GKKRVRWADLE eluted bands were excised peptide is an excellent tool to displace PP1 regulatory sub- from the gel and proteins identified by mass spectrometry units, we tried several peptide concentrations to optimize [Additional file 2]. Many previously characterized and displacement from the matrix (Figure 1e). This demon- new potential PP1 binding proteins were identified. To strated that 0.1 mM peptide was sufficient to displace ZAP validate our findings in different organisms, we then and p99, but 0.5 to 2 mM peptide was needed to remove switched our attention to the human HeLa cell line. Here all of the more abundant and perhaps higher affinity the GKKRARAADLE, GKKRVRWADLE and NaSCN col- binding NIPP1 protein. At the 2 mM peptide concentra- umn elutions again show distinct protein patterns (Fig tion we noted that displacement was in fact not as effec- 2a). Mass spectrometry identification of individual bands tive for ZAP or p99. In subsequent experiments we eluted from the GKKRVRWADLE displacement is shown in by first incubating with 0.5 mM, then 2 mM peptide to Additional file 3 and file 4. For many of the identified pro- ensure all PP1 complex proteins were displaced. To ensure teins we were able to obtain antibodies and blotting of the the specificity of the method, we included 2 independent individual elution steps confirmed that the GKKRARAA- negative controls. Nuclear phosphatases bound to the DLE peptide did not elute any of the known or putative microcystin matrix were equally divided into 3 parts and new PP1 interactors, supporting the idea that all of the eluted with either the GKKRVRWADLE peptide, the GKKRVRWADLE released proteins reside in PP1 com- GKKRVRWADLE peptide where the key PP1 interacting plexes and that many are direct PP1 binding proteins (Fig- amino acids (V and W) were replaced by alanine (GKKRA- ure 2b). Of note is the abundant presence of putative PP1 RAADLE), or a scrambled ZAP peptide (KLRGEVAK- docking sites within the sequences of identified proteins DWR). These fractions were blotted for ZAP, p99 and (additional files 2 and 4). Although we predict only PP1 NIPP1 (Figure 1f) confirming the specificity of the dis- binding/complex proteins should elute with docking site placement technique. peptide, mass spectrometry reveals that some PP1 cata- lytic and PP2A components did release with GKKRVRWA- To extend this method to PP1 complexes outside the DLE elution, while blotting showed that the vast majority nucleus, we isolated glycogen particles from skeletal mus- of the total pool of these subunits were retained on the cle where the well characterized PP1 regulatory subunit matrix (Figure 2b). This is not surprising given that some GM resides and functions to target PP1 to glycogen [22]. As complexes likely contain both PP1 and PP2A [33] and shown in Figure 1g, the peptide GKKRVRWADLE effec- other proteins may contain more than one PP1 dock site. tively displaced GM from the microcystin-matrix. In addi- The lack of displacement with the GKKRARAADLE pep- tion, prior to the RVRW displacement, we performed an tide, and subsequent elution with GKKRVRWADLE pro- incubation with the GKKRARAADLE peptide and, as pre- vide compelling evidence that PP1 directly docks or is in dicted, due to the absence of the key V and W residues in these protein complexes. To explore this further we per- the docking motif, this peptide did not displace the GM formed immunoprecipitations with antibodies we protein from the matrix which supports the idea that the obtained for many of these proteins and in every instance peptide elution is specific for PP1 complexes. All subse- tested we could demonstrate the interaction with PP1 quent experiments included the GKKRARAADLE peptide (Figure 3). These data again support the idea that we spe- elution and buffer wash step prior to GKKRVRWADLE cifically eluted PP1 complex proteins from the matrix. peptide displacement. One reason for using the RVRW peptide elution was to specifically displace proteins and not release any polypep- Having developed this methodology, we scaled it up to tides that associate non-specifically. It is known that some specifically release PP1 interacting proteins from the proteins bind non-specifically to Sepharose beads and matrix with the aim of identifying novel, less abundant thus we sought to determine if the proteins discovered nuclear PP1 regulatory subunits and/or proteins that here were retained on Sepharose alone and eluted with reside in PP1 complexes. We predict many more unrecog- the GKKRVRWADLE peptide. Therefore, as an additional nized PP1 regulatory subunits exist whose identification control, we passed a HeLa nuclear extract over Sepharose may have been masked in previous studies when all beads alone, washed as usual, then eluted with the microcystin-sensitive phosphatases (especially the abun- GKKRARAADLE and GKKRVRWADLE peptides and dant PP2A complex) and non-specifically bound proteins NaSCN, as described for the microcystin-matrix. These were released from the matrix using NaSCN alone [31]. samples were then western blotted for a selection of inter-

Page 4 of 10 (page number not for citation purposes) BMC Biochemistry 2008, 9:28 http://www.biomedcentral.com/1471-2091/9/28

DisplacementFigure 2 affinity chromatography purification of PP1 complexes from HeLa cell nuclei Displacement affinity chromatography purification of PP1 complexes from HeLa cell nuclei. (a) A HeLa cell nuclear extract was incubated with microcystin-Sepharose, washed extensively and eluted with the GKKRARAADLE peptide, followed by GKKRVRWADLE peptide and finally with 3 M NaSCN. Each fraction was concentrated to an equal volume, run on 4–12% SDS-PAGE (Invitrogen) and stained with Collodial blue. Individual bands were excised, trypsin digested and identified by mass spectrometry (see Additional files 2 and 4). In (b) the same samples from panel (a) were blotted to a membrane and probed with antibodies to proteins identified by mass spectrometry. Protein names are defined in abbreviations list.

Page 5 of 10 (page number not for citation purposes) BMC Biochemistry 2008, 9:28 http://www.biomedcentral.com/1471-2091/9/28

Co-immunprecipitationingFigure and complex3 proteins of PP1 with HeLa nuclear PP1 bind- Co-immunprecipitation of PP1 with HeLa nuclear PP1 binding and complex proteins. Antibodies were obtained for the proteins shown and immunoprecipitations performed. Immunoprecipitated proteins were run on SDS- PAGE, transferred to a membrane and probed with the same antibodies as indicated plus a pan PP1 antibody to determine if PP1 co-immunoprecipitates. The immunoprecipitating anti- body is shown on the left and in all cases a pre-immune serum control was done in parallel (PIS). The blotting anti- body it indicated on the right.

actors that were previously retained on the microcystin- Sepharose and eluted with the GKKRVRWADLE peptide (Figure 2b). When the entire GKKRARAADLE, GKKRVR- WADLE and NaSCN fractions were individually concen- trated and run on a single lane and blotted versus 1/175 of the input (which gave a strong western signal) we did weakly detect several of these proteins in the NaSCN elu- tion (as expected), but only noted signals for filamin A and PSF in the GKKRVRWADLE peptide elution (data not shown). Given that this is in comparison to 1/175 of the input, this represents a very minor background displace- ment and functions to remind us that, as in any explora- tory work, additional experiments are needed to confirm true interactors. Interestingly, PSF has previously been defined as a PP1 docking protein [34].

It is notable from the mass spectrometry data that we identified many nuclear processes or complexes not previ- ously linked to PP1. For instance, we found numerous TIP60 (TRRAP) complex proteins [35]. Immunoprecipita- tion of the TIP60 components TRRAP or p400 (Domino) were able to bring down PP1, confirming that PP1 is a pre- viously unrecognized component of this complex. Puta- tive interacting proteins or complexes identified here will need to be confirmed by co-immunoprecipitation experi- ments or other methods, and then ultimately all com- plexes, including the TRRAP complex, will need to be studied to define the role of PP1 binding to that protein or group of proteins.

A comparison of the proteins eluted and identified from rat and human nuclei reveals that most of the proteins are the same (although the number of rat proteins is greater), but several differences are apparent. First, some differ- ences are not surprising given that the HeLa cell line expresses numerous proteins that differ from the rat liver cell proteome. In addition, there are quite a number of proteins in the rat samples that appear in multiple bands excised from the gel, but this only occurs a few times in Figure 3 the HeLa bands. This is likely due to the fact that prepara- tion of the rat nuclei from liver tissue is more time con-

Page 6 of 10 (page number not for citation purposes) BMC Biochemistry 2008, 9:28 http://www.biomedcentral.com/1471-2091/9/28

suming and thus we expect increased proteolysis of rose or control matrix for 1 h at 4°C and then washed in proteins and therefore many more hits of the same pro- a column with 250 mL of buffer A (25 mM Tris-HCl pH tein in bands of many masses. Comparing rat liver versus 7.5, 0.1 mM EGTA, 1 mM benzamidine, 0.1 mM PMSF, cells grown in culture also means that we started with 0.1% β-ME) plus 300 mM NaCl. All peptides were dis- much more protein from the rat preparation and this cer- solved in buffer A plus 300 mM NaCl and the pH adjusted tainly contributed to the identification of more lower to 7.5 with 1 N NaOH by spotting on pH paper. The abundance proteins. matrix (1 mL) was first incubated with 2 mL 0.5 mM pep- tide for 30 min. and the eluted protein collected with the To date several approaches have been used to define the addition of 2 mL buffer A plus 300 mM NaCl. The matrix PP1-interactome of the cell, including total (NaSCN) elu- was then incubated with 2 mL 2 mM peptide for 30 min. tion from microcystin-Sepharose followed by overlay and the eluted protein collected with the addition of 5 mL analysis, SILAC mass spectrometry based PP1 binding buffer A plus 300 mM NaCl. After washing with an addi- partner investigation and most recently an antibody array tional 30 mL of buffer A plus 300 mM NaCl, the column approach for selected putative targets [11,31,36,37]. Any was eluted with 3 M sodium isothiocyanate in buffer A methodology has its limitations; our work certainly repre- [9]. The peptides used were RPKRKRKNSRVTFSEDDEII sents a subgroup of the total PP1 targeted protein/com- (from human NIPP1), GKKRVRWADLE (from human plexes of the nucleus for reasons discussed above and ZAP), GKKRARAADLE (from human ZAP with alanine because our nuclear extraction procedure likely only substitutions for V and W) and KLRGEVAKDWR (scram- released some fraction of the total PP1 complexes. It is bled GKKRVRWADLE). The 0.5 and 2 mM peptide elu- clear that many more PP1 binding partners exist than orig- tions were pooled and immediately concentrated in a inally thought and it will be the combination of multiple centriprep 10, then centricon 10 to 40 μL and boiled in 5× approaches that will ultimately define the plethora of SDS-cocktail. The NaSCN elution was dialyzed extensively roles for PP1 in the nucleus and other cellular compart- and concentrated to 40 μL as described for the peptide ments. elutions and boiled in 5× SDS-cocktail. Samples were run on 10% SDS-PAGE or in some cases on 4–12% gradient Conclusion gels (Invitrogen). We have developed a method to specifically elute proteins that reside in PP1 complexes or bind directly to PP1 after Isolation of rabbit skeletal muscle glycogen particles and retention on the affinity matrix microcystin-Sepharose. chromatography on MC-Sepharose This approach should be valuable to isolate new PP1 Skeletal muscle from the back and hind legs of one rabbit binding proteins and to link PP1 function to as yet unrec- was removed, placed on ice, minced and homogenized ognized PP1 regulated cellular processes across a broad with 2.5 L/kg homogenization buffer (2 mM EDTA, 2 mM range of tissues, cell types and organisms. This technique EGTA, 0.1% B-ME, 1 mM benzamidine, 0.1 mM PMSF) in also lends further credibility to the notion that it could be a blender and centrifuged in a Beckman J6 at 4200 rpm for possible to develop peptides as tools for the specific dis- 30 min. at 4°C. The supernatant was decanted through ruption of protein protein interactions with potential glass wool in a funnel and the pH of the solution lowered usage for therapeutic purposes. to 6.1 with 1 M acetic acid. After 15 min on ice, glycogen was pelleted by centrifugation at 4200 rpm for 30 min. at Methods 4°C. The glycogen pellet was resuspend in 100 mL of the HeLa nuclear extracts following buffer (50 mM Tris-HCl pH 7, 2 mM EGTA, 5% For HeLa cells, nuclei were isolated and extracted from ~1 V/V glycerol, 4 μg/mL leupeptin, 1 mM benzamidine, 0.1 × 109 cells grown in spinner flasks using the method out- mM PMSF, and 0.1% β-ME.) and loaded on a 1 mL micro- lined at http://www.lamondlab.com. Sonicated, extracted cystin-Sepharose column equilibrated with microcystin- nuclei were made to 0.42 M NaCl, mixed end-over-end for Sepharose buffer A. After incubation for 1.5 hour with an 10 min. at 4°C and clarified by centrifugation at 10,000 g end-over-end machine, then washing with microcystin- for 10 min. This clarified extract was then mixed with the Sepharose buffer A plus 0.3 M NaCl (200 column vol- microcystin matrix. umes or until no protein comes off in a Bradford assay) the matrix was eluted with peptides and NaSCN as Affinity displacement chromatography described above. Microcystin-Sepharose was prepared as described [7,9]. Control matrix was prepared by coupling to Tris instead of Mass spectrometry aminoethanethiol-microcystin. In a typical nuclear prepa- For protein identification by mass spectrometry bands ration, the livers of 10 rats were homogenized, nuclei iso- were excised from colloidal blue stained gels, digested lated and proteins extracted [31]. Nuclear proteins were with trypsin and LC-MS/MS was performed as in Ulke- incubated end over end with 1 mL of microcystin-Sepha- Lemée et al [32]. Peak lists were searched with Mudpit

Page 7 of 10 (page number not for citation purposes) BMC Biochemistry 2008, 9:28 http://www.biomedcentral.com/1471-2091/9/28

scoring using Mascot version (v2.2). We used the criteria LRR-Ri: Leucine-rich repeats (LRRs): ribonuclease inhibi- of 2 matching peptides for a positive identification. tor; GRP78: Glucose-Regulated Protein of 78-kDa.

Immunoprecipitations Authors' contributions Antibodies to proteins or an equivalent amount of pre- GBGM, LTM and AIL conceived of the study, and partici- immune serum purified IgG were coupled to Protein A- pated in its design. GBGM, LTM, VD and MN carried out Sepharose (GE Healthcare) and incubated with HeLa cell the biochemical studies and drafted the manuscript. DGC, extracts (from 4 × 10 cm plates) end over end for 2 h at RG and YWL carried out the mass spectrometry to identify 4°C. After sedimentation, IP pellets were washed 2× with proteins. PBS, 2× with PBS plus 0.05% NP-40 and 150 mM NaCl and 2 more times in PBS, before boiling in SDS-cocktail. Additional material In all cases, antibody and pre-immune serum IPs were incubated with identical amounts of protein, washed in parallel and eluted with an equal volume of cocktail to Additional file 1 Supplementary Figure 1. Identification of novel rat liver nuclear PP1 allow direct comparison on western blots. binding and complex proteins by displacement affinity chromatogra- phy. Protein was extracted from isolated rat liver nuclei, incubated with Antibodies were kindly provided by individuals or pur- microcystin-Sepharose, the matrix washed extensively and eluted with chased as indicated in brackets. PP1, NIPP1, p99 GKKRVRWADLE peptide, followed by elution with 3 M NaSCN [31]. (PNUTS), SAP155, ZAP and their use are detailed in Tran GKKRVRWADLE and NaSCN eluted samples were concentrated sepa- et al [31]. Other antibodies were PR65 (B. Hemmings), rately to an equal volume and run on 10% SDS-PAGE and stained with Collodial blue. In a parallel experiment, the bands shown above were GM and M110 (P. Cohen), DPK-1 (SP Lees-Miller), excised, trypsin digested and analyzed by mass spectrometry for identifica- NUP153 (M. Lohka), DDX5 (F. Fuller-Pace), RAVER-1 (B. tion. The top matched identified proteins for each band(s) are indicated Jockusch), TOPOIIα (E. Kurz), TRRAP, Envoplakin, to the left of the figure. Additional matches for each excised band and HDAC6, PSF (Santa Cruz), Ki-67, p53bp2 [ASPP2], details of protein identifications are in Additional file 2 online. p54nrb/nono (BD Biosciences), Rif-1, p84 [Thoc1], P400 Click here for file (Abcam), Filamin-A (Chemicon), NUMA-1 (Nova Biolog- [http://www.biomedcentral.com/content/supplementary/1471- 2091-9-28-S1.jpeg] icals), DDX18, DDX17 [DNA helicaseA] (Bethyl Lab), HDAC1 (Cell Signaling), PP2A-C (Transduction Labs). Additional file 2 Additional file 2-Supplementary Table 1. Identification of proteins Abbreviations eluted from microcystin-Sepharose using an 'RVRW' peptide (Rattus DPK1: DNA-dependent protein kinase; TRRAP: transfor- norvegicus). Proteins were run on SDS-PAGE and stained with Collodial mation/transcription domain-associated protein; RIF1: blue. Individual bands were excised, digested with trypsin and proteins -interacting factor 1; NUMA-1: nuclear mitotic appa- identified by Mass Spectrometry. These proteins are shown by gel band excised with and protein name/description, gene accession numbers, gene ratus protein 1; TOPOIIα: topoisomerase IIα; SAP155: name and whether they were previously identified has a PP1 targeting spliceosome-associated protein 155; NUP153: nucleop- subunit (Known PP1 TS). orin 153; M110/MYPT1: targeting subunit 110; Click here for file p53bp2: p53 binding protein 2; PSF: PTB-associated splic- [http://www.biomedcentral.com/content/supplementary/1471- ing factor; RAVER1: ribonucleoprotein: PTB-binding 1; 2091-9-28-S2.xls] PR65: PP2A scaffolding or A subunit; p400/Domino: E1A-binding protein p400; ASPP: apoptosis stimulating Additional file 3 Supplementary Figure 2. Identification of HeLa nuclear PP1 binding proteins of p53; NaSCN: sodium isothiocyanate; HDAC: and complex proteins by displacement affinity chromatography. The histone deacetylase; DDX: DEAD box family of RNA heli- GKKRVRWADLE elution from Figure 2a has been cropped and the top cases; PP1: protein phosphatase one; PP2A: protein phos- match identified proteins for each band(s) are indicated on the figure. phatase 2A; LCP1: epidermal Langerhans cell protein; Additional matches for each excised band and details of protein identifi- KPI-2: Kinase/Phosphatase/Inhibitor-2; NIPP: nuclear cations are in Additional file 4 (Supplementary Table 2) online. inhibitor of protein phosphatase-1: ZAP (ZAP3): YLPM- Click here for file [http://www.biomedcentral.com/content/supplementary/1471- motif contain PP1 interactor; p99: PP1 interactor (also 2091-9-28-S3.jpeg] PNUTS); GM: skeletal muscle glycogen binding PP1 inter- actor; MC: microcystin; p84/Thoc1: THO complex 1; p54nrb/nono: 54 kDa nuclear RNA binding protein; NUP50: nucleoporin 50; TIP49a/RUVBL2: TBP-interact- ing protein 49/RuvB-like 2; PP4: protein phosphatase four; Repo-Man: recruits PP1 onto mitotic chromatin at anaphase; MBS85: myosin-binding subunit of 85 kDa;

Page 8 of 10 (page number not for citation purposes) BMC Biochemistry 2008, 9:28 http://www.biomedcentral.com/1471-2091/9/28

15. Terrak M, Kerff F, Langsetmo K, Tao T, Dominguez R: Structural Additional file 4 basis of protein phosphatase 1 regulation. Nature 2004, 429(6993):780-784. Supplementary Table 2. Identification of proteins eluted from micro- 16. Neduva V, Linding R, Su-Angrand I, Stark A, de Masi F, Gibson TJ, cystin-Sepharose using an 'RVRW' peptide (Homo Sapiens). Proteins Lewis J, Serrano L, Russell RB: Systematic discovery of new rec- were run on SDS-PAGE and stained with Collodial blue. Individual ognition peptides mediating protein interaction networks. bands were excised, digested with trypsin and proteins identified by Mass PLoS Biol 2005, 3(12):e405. Spectrometry. These proteins are shown by gel band excised with gene 17. Hurley TD, Yang J, Zhang L, Goodwin KD, Zou Q, Cortese M, Dun- ker AK, DePaoli-Roach AA: Structural basis for regulation of accession numbers, gene name, number of peptide identified by mass spec- protein phosphatase 1 by inhibitor-2. J Biol Chem 2007, trometry and whether they were previously identified has a PP1 targeting 282(39):28874-28883. subunit (Known PP1 TS). 18. Ayllon V, Cayla X, Garcia A, Fleischer A, Rebollo A: The anti-apop- Click here for file totic molecules Bcl-xL and Bcl-w target protein phosphatase [http://www.biomedcentral.com/content/supplementary/1471- 1alpha to Bad. Eur J Immunol 2002, 32(7):1847-1855. 2091-9-28-S4.xls] 19. Chang JS, Henry K, Wolf BL, Geli M, Lemmon SK: Protein phos- phatase-1 binding to scd5p is important for regulation of actin organization and endocytosis in yeast. J Biol Chem 2002, 277(50):48002-48008. 20. Carmody LCBAn, Bass MA, Colbran RJ: Selective targeting of the gamma1 isoform of protein phosphatase 1 to F-actin in Acknowledgements intact cells requires multiple domains in spinophilin and neu- This work was supported by the Natural Sciences and Engineering rabin. FASEB J 2008, 6(6):1660-1671. Research Council of Canada and the Alberta Cancer Board. 21. Ceulemans H, Bollen M: Functional diversity of protein phos- phatase-1, a cellular economizer and reset button. Physiol Rev 2004, 84(1):1-39. References 22. Cohen PT: Protein phosphatase 1 – targeted in many direc- 1. Olsen JV, Blagoev B, Gnad F, Macek B, Kumar C, Mortensen P, Mann tions. J Cell Sci 2002, 115(Pt 2):241-256. M: Global, In Vivo, and Site-Specific Phosphorylation Dynam- 23. Meiselbach H, Sticht H, Enz R: Structural analysis of the protein ics in Signaling Networks. Cell 2006, 127(3):635-648. phosphatase 1 docking motif: molecular description of bind- 2. Cohen P: The origins of protein phosphorylation. Nat Cell Biol ing specificities identifies interacting proteins. Chem Biol 2006, 2002, 4(5):E127-130. 13(1):49-59. 3. Moorhead GB, Trinkle-Mulcahy L, Ulke-Lemée A: Emerging roles 24. Zhao S, Lee EY: A protein phosphatase-1-binding motif identi- of nuclear protein phosphatases. Nat Rev Mol Cell Biol 2007, fied by the panning of a random peptide display library. J Biol 8(3):234-244. Chem 1997, 272(45):28368-28372. 4. Tonks NK: Protein tyrosine phosphatases: from genes, to 25. Beullens M, Van Eynde A, Vulsteke V, Connor J, Shenolikar S, Stal- function, to disease. Nat Rev Mol Cell Biol 2006, 7(11):833-846. mans W, Bollen M: Molecular determinants of nuclear protein 5. Kerk D, Templeton G, Moorhead GB: Evolutionary radiation pat- phosphatase-1 regulation by NIPP-1. J Biol Chem 1999, tern of novel protein phosphatases revealed by analysis of 274(20):14053-14061. protein data from the completely sequenced genomes of 26. Johnson DF, Moorhead G, Caudwell FB, Cohen P, Chen YH, Chen humans, green algae, and higher plants. Plant Physiol 2008, MX, Cohen PT: Identification of protein-phosphatase-1-bind- 146(2):351-367. ing domains on the glycogen and myofibrillar targetting sub- 6. MacKintosh RW, Dalby KN, Campbell DG, Cohen PT, Cohen P, units. Eur J Biochem 1996, 239(2):317-325. MacKintosh C: The cyanobacterial toxin microcystin binds 27. Piljic A, Schultz C: Simultaneous recording of multiple cellular covalently to cysteine-273 on protein phosphatase 1. FEBS events by FRET. ACS Chem Biol 2008, 3(3):156-160. Lett 1995, 371(3):236-240. 28. Guergnon J, Dessauge F, Dominguez V, Viallet J, Bonnefoy S, Yuste VJ, 7. Moorhead GB, Haystead TA, MacKintosh C: Synthesis and use of Mercereau-Puijalon O, Cayla X, Rebollo A, Susin SA, et al.: Use of the protein phosphatase affinity matrices microcystin- penetrating peptides interacting with PP1/PP2A proteins as sepharose and microcystin-biotin-sepharose. Methods Mol Biol a general approach for a drug phosphatase technology. Mol 2007, 365:39-45. Pharmacol 2006, 69(4):1115-1124. 8. Nishiwaki S, Fujiki H, Suganuma M, Nishiwaki-Matsushima R, Sugimura 29. Alto NM, Soderling SH, Hoshi N, Langeberg LK, Fayos R, Jennings PA, T: Rapid purification of protein phosphatase 2A from mouse Scott JD: Bioinformatic design of A-kinase anchoring protein- brain by microcystin-affinity chromatography. FEBS Lett 1991, in silico: a potent and selective peptide antagonist of type II 279(1):115-118. protein kinase A anchoring. Proc Natl Acad Sci USA 2003, 9. Moorhead G, MacKintosh RW, Morrice N, Gallagher T, MacKintosh 100(8):4445-4450. C: Purification of type 1 protein (serine/threonine) phos- 30. Trinkle-Mulcahy L, Ajuh P, Prescott A, Claverie-Martin F, Cohen S, phatases by microcystin-Sepharose affinity chromatogra- Lamond AI, Cohen P: Nuclear organisation of NIPP1, a regula- phy. FEBS Lett 1994, 356(1):46-50. tory subunit of protein phosphatase 1 that associates with 10. Moorhead G, MacKintosh C, Morrice N, Cohen P: Purification of pre-mRNA splicing factors. J Cell Sci 1999, 112(Pt 2):157-168. the hepatic glycogen-associated form of protein phos- 31. Tran HT, Ulke A, Morrice N, Johannes CJ, Moorhead GB: Pro- phatase-1 by microcystin-Sepharose affinity chromatogra- teomic characterization of protein phosphatase complexes phy. FEBS Lett 1995, 362(2):101-105. of the mammalian nucleus. Mol Cell Proteomics 2004, 11. Campos M, Fadden P, Alms G, Qian Z, Haystead TA: Identification 3(3):257-265. of protein phosphatase-1-binding proteins by microcystin- 32. Ulke-Lemee A, Trinkle-Mulcahy L, Chaulk S, Bernstein NK, Morrice biotin affinity chromatography. J Biol Chem 1996, N, Glover M, Lamond AI, Moorhead GB: The nuclear PP1 inter- 271(45):28478-28484. acting protein ZAP3 (ZAP) is a putative nucleoside kinase 12. Brush MH, Guardiola A, Connor JH, Yao TP, Shenolikar S: Deacty- that complexes with SAM68, CIA, NF110/45, and HNRNP- lase inhibitors disrupt cellular complexes containing protein G. Biochim Biophys Acta 2007, 1774(10):1339-1350. phosphatases and deacetylases. J Biol Chem 2004, 33. Shi Y, Reddy B, Manley JL: PP1/PP2A phosphatases are required 279(9):7685-7691. for the second step of Pre-mRNA splicing and target specific 13. Bollen M: Combinatorial control of protein phosphatase-1. snRNP proteins. Mol Cell 2006, 23(6):819-829. Trends Biochem Sci 2001, 26(7):426-431. 34. Hirano K, Erdödi F, Patton JG, Hartshorne DJ: Interaction of pro- 14. Egloff MP, Johnson DF, Moorhead G, Cohen PT, Cohen P, Barford D: tein phosphatase type 1 with a splicing factor. FEBS Letters Structural basis for the recognition of regulatory subunits by 1996, 389(2):191. the catalytic subunit of protein phosphatase 1. Embo J 1997, 35. Lee KK, Workman JL: Histone acetyltransferase complexes: 16(8):1876-1887. one size doesn't fit all. Nat Rev Mol Cell Biol 2007, 8(4):284-295.

Page 9 of 10 (page number not for citation purposes) BMC Biochemistry 2008, 9:28 http://www.biomedcentral.com/1471-2091/9/28

36. Flores-Delgado G, Liu CW, Sposto R, Berndt N: A limited screen for protein interactions reveals new roles for protein phos- phatase 1 in cell cycle control and apoptosis. J Proteome Res 2007, 6(3):1165-1175. 37. Trinkle-Mulcahy L, Andersen J, Lam YW, Moorhead G, Mann M, Lam- ond AI: Repo-Man recruits PP1 gamma to chromatin and is essential for cell viability. J Cell Biol 2006, 172(5):679-692.

Publish with BioMed Central and every scientist can read your work free of charge "BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime." Sir Paul Nurse, Cancer Research UK Your research papers will be: available free of charge to the entire biomedical community peer reviewed and published immediately upon acceptance cited in PubMed and archived on PubMed Central yours — you keep the copyright

Submit your manuscript here: BioMedcentral http://www.biomedcentral.com/info/publishing_adv.asp

Page 10 of 10 (page number not for citation purposes)